10TH INTERNATIONAL SYMPOSIUM ON PARTICLE IMAGE VELOCIMETRY – PIV13 Delft, The Netherlands, July 1-3, 2013
Flow characteristics of ionic liquid–aqueous two-phase flows in small
channels
Dimitrios Tsaoulidis, Qi Li and Panagiota Angeli
Department of Chemical Engineering, University College London, Torrington Place, WC1E 7JE, UK
dimitrios.tsaoulidis.10@ucl.ac.uk, corresponding author: p.angeli@ucl.ac.uk
ABSTRACT
The hydrodynamic characteristics of liquid-liquid segmented flow in small channels were studied using bright field μ-PIV. Experiments were carried out in a 0.5 mm ID channel made of PFA and flow parameters key for mass transfer applications, such as film thickness, circulation patterns and plug length were measured. The two test liquids used were a nitric acid solution of 3M and TBP/ionic liquid (30% v/v) mixtures. Three different ionic liquids were used, while the aqueous phase formed the dispersed plugs at all experimental conditions. It was found that the plug length decreased linearly as the mixture velocity (Vmix) increased. A thin film was formed at all experimental conditions, and it increased
with increasing mixture velocity. For the more viscous ionic liquids it was found that at constant mixture velocity the film thickness increased with the velocity of the carrier phase. The circulation patterns within the plugs were affected by the plug speed and the plug length.
1. INTRODUCTION
Liquid-liquid two phase systems have found numerous applications in many chemical processes, such as solvent extraction, emulsification, phase transfer catalysis, and polymerization [1]. Processes in intensified small scale units offer among others, higher efficiency, reduction of waste and safety. Despite the absence of turbulence, mixing and heat transfer can be significantly enhanced because of the thin fluid layers that are established in the small fluidic channels [2-5]. There are relatively fewer studies that focus on the hydrodynamics and mixing characteristics in liquid-liquid flows in microchannels compared to gas-liquid flows [6]. However, liquid-liquid systems can be more challenging, since, depending on the wetting properties of the two liquids with the channel material, either phase can preferentially wet the channel wall and become continuous. The most common flow configurations that have been studied are segmented (plug) and parallel flow. In segmented (plug) flow, one liquid flows within the other in the form of droplets that have diameter larger than the channel diameter. This flow configuration offers excellent mass transfer and a well-defined interfacial area. The presence of a thin wall film that surrounds the plugs increases the specific interfacial area for mass transfer, compared to cases where no film is present [7]. In addition, circulation patterns appear within the plugs which further enhance the mass transfer; these patterns will also be affected by the presence of the film. Despite its significance, film thickness has not been well investigated for liquid-liquid segmented flow systems in small scale channels.
Understanding the mixing characteristics and the hydrodynamics of two phase flows in small scale devices is crucial for considering potential industrial applications. Important mixing characteristics such as circulation time in segmented flows can be derived from the velocity fields, location of the stagnation points and the vortex cores [8]. Micro-Particle Image Velocimetry (μ-PIV) can be used to extract multipoint information of the velocity inside a single plug or slug with high accuracy and spatial resolution and in a non-intrusive way [9]. There are a number of investigations in gas-liquid systems that involve the application of μ-PIV, but fewer in the case of gas-liquid-gas-liquid systems [10-12].
In the current study, the hydrodynamic properties and mixing characteristics of an ionic liquid-aqueous segmented two phase flow in small channels are studied. This investigation is complementary to an ongoing research on the use of ionic liquids for spent nuclear fuel reprocessing in small scale separators. Ionic liquids are salts composed entirely of ions that have low melting points (below 100 ºC), while many of them are liquids at room temperature [13]. Their properties can be tuned by the choice of the anion and cation depending on the application, and a large number of combinations can be achieved. In recent years ionic liquids have applied in many catalytic reactions, as well as in synthesis and separation processes [14].
2.1. Materials
The ionic liquids used in this work are 1-butyl-3-methylimidazolium bis{(trifluoromethyl)sulfonyl}amide [C4mim][NTf2], 1-decyl-3-methylimidazolium bis{(trifluoromethyl)sulfonyl}amide [C10mim][NTf2], and
trihexyltetradecylphosphonium bis{(trifluoromethyl)sulfonyl}amide [P66614][NTf2]. They were produced at QUILL
research centre, Queen’s University of Belfast, and their properties are summarized in Table 1. Although these ionic liquids are hydrophobic, they still absorb small amounts of water depending on the nitric acid concentration in the aqueous phase. The absorbed water affects their viscosity. The viscosities of the ionic liquids saturated with water were measured using a digital Rheometer DV-III Ultra (Brookfield) and were found to decrease by 15-20% compared to the values of pure ionic liquids. The tri-n-butylphosphate (TBP) used was obtained from Sigma-Aldrich, while the HNO3
aqueous solutions (3M) were prepared in the UCL laboratory from HNO3 (65%) solution of general purpose grade by
Fisher Scientific.
Table 1: Properties of Ionic Liquids. 1 Properties of pure IL as measured at QUILL research centre. 2 Interfacial tension between TBP/IL system and the aqueous phase as measured at the UCL laboratory. 3 Surface tension of the TBP/IL system as measured at the UCL laboratory.
Properties (at room temperature) [C4mim][NTf2] [C10mim][NTf2] [P66614][NTf2]
1Viscosity, μ / kg m-1 s-1 0.052 0.124 0.296
1Density, ρ / kg m-3 1420 1260 1065
2Interfacial tension, γ / N m-1 (∙10-3) 10.01 11.51 11.92
3Surface tension, σ / N m-1 (∙10-3) 29.31 28.32 28.53
2.2. Experimental apparatus & procedure
Experiments were carried out in the experimental setup shown in figure 1. The two fluids were fed into the test channel made of Teflon by using a double high-precision pump (KdScientific). A T-junction made of Teflon with all the branches having the same internal diameter as the test channel, was used to bring the fluids together. The experiments were carried out at total volumetric flow rates varying from 7.1 cm3 h-1 to 42.4 cm3 h-1. Plug flow was obtained at all experimental conditions with the aqueous phase flowing as non-continuous plugs (dispersed phase) within the TBP/IL (30% v/v) phase (continuous phase).
Flow characteristics in the two-phase system [3M nitric acid solution and TBP/IL (30% v/v) mixture] were studied with micro-Particle Image Velocimetry. For the measurements the non-conventional bright field illumination approach was used, as it is easy to implement and it allows velocity to be measured at the same time as mass transfer. In this the aqueous phase was seeded with polystyrene particles (3 μm) and the test section was backlit using a 60 Watt continuous arc lamp. Images of dark particles in a bright background were obtained. A high-speed camera (Photron Fastcam-ultima APX) equipped with a 5X magnification lens acquired 2MP 10-bit images at 2000 Hz. To accurately locate the observation plane at channel mid-plane, both the camera and the test section were mounted in micrometer stages and traverses, which allowed a 3D relative motion between them. In addition, the test section was enclosed in a flat visualisation box filled with water to minimise reflections at the interface.
The time interval between an image pair was chosen depending on the expected velocity inside the channel and the required spatial resolution. In the current study a square discretization domain of 32 x 32 pixels with 50% overlap was used, which provided a spatial resolution of up to 28 μm along the x and y axes, while the effective depth of correlation (2z) was expected to be of the order of 100 μm.
The TSI PIV platform Insight 3G was used to calculate velocity vectors in the aqueous plugs. The circulation patterns were plotted in Tecplot. For the calculation of the film thickness (δ) and the plug/slug size, a post-processing routine was developed to detect the plug interface. Image binarization was carried out by setting a threshold pixel value to discriminate the aqueous medium from the TBP/IL phase.
Figure 1. Schematic representation of the brightfield experimental set up.
3. RESULTS AND DISCUSION 3.1. Plug length
The plug length (Lp) was measured for the different experimental conditions by using the image binarization technique
discussed in section 2.2. It can be seen from figure 2 that the plug length decreased linearly as the mixture velocity (Vmix) increased. This decrease is attributed to the rapid penetration of one phase into the other at high velocities. In
addition, the shape of the plug was found to change as a function of the mixture velocity (Vmix). In particular, at low
mixture velocities the shape of the plug was symmetric and both ends had a convex shape, whilst at high mixture velocities the plug acquired a “bullet” shape, where the rear cup became flatter and the front sharper.
At all experimental conditions studied there was a film present between the plug and the channel wall and increases the specific interfacial area available for mass transfer. The film thickness (δ) between the plug and the channel wall was studied for different properties of the continuous phase, i.e. different ionic liquids, and different mixture velocities. The post-processing routine of interface detection was applied, and the width of the plugs was calculated. Then the film thickness was derived by subtracting the width of the plug from the internal diameter of the channel and dividing by 2. The film thickness is plotted against the continuous phase Reynolds number, Rec, in figure 3 when the [C4mim][NTf2]
was used as continuous phase and at flow rate ratio (QIL/QHNO3) equal to 1. As can be seen, the dimensional film
thickness (δ) increased asymptotically with continuous phase Reynold number, Rec, following a logarithmic trend of
R2=0.9913. The Rec was calculated as follows:
Rec= Vcρcd/μc
where Vc is the superficial velocity of the continuous phase, ρc is the density of the continuous phase, d is the internal
diameter of the small channel and μc is the viscosity of the continuous phase. A similar trend was also found for the
other ionic liquids.
Figure 3. Dimensional film thickness (δ) as a function of the Reynolds number of the continuous phase (Rec) for
TBP/[C4mim]NTf2] (30% v/v) as continuous phase and flow rate ratio equal to 1.
The film thickness at a constant mixture velocity of 0.02 m s-1 is shown in figure 4 against the continuous phase velocity Vc for all the ILs studied. In this case, the flow rate ratio (QIL/QHNO3) also changes from 0.5 to 1 as the continuous phase
velocity changes from 0.066 m s-1 to 0.01 m s-1 respectively. It can be seen that as the velocity of the continuous phase increases, the film thickness also increases. This increase is more obvious in the case of the more viscous ILs ([C10mim]NTf2] and [P66614][NTf2]), whilst in the case of [C4mim]NTf2] the film thickness remains almost constant. At
Figure 4. Dimensional film thickness as a function of the superficial velocity of the continuous phase (Vc) for the 3
different TBP/IL (30% v/v) mixtures at a constant mixture velocity (Vmix) of 0.02 m s-1.
3.3. Circulation patterns
Velocity profiles were acquired at all experimental conditions in the aqueous plugs. To obtain circulation patterns velocities relative to the plug speed have to be calculated within the plugs. The speed of the plug was found by averaging the velocities in the plug. The relative velocities for flow rate ratio (QIL/QHNO3) 0.5 and a constant mixture
velocity of 0.02 m s-1, are shown in Figure 5 for the three different IL mixtures.
It can be observed from figure 5 that as the viscosity of the carrier phase increases (AC), the relative velocities decrease, implying that the circulation within the aqueous plug becomes less intense. This is attributed to the decrease in plug speed with the increase in viscosity of the continuous phase.
Figure 5. Velocity in the aqueous phase relative to the average plug speed for the three different TPB/IL (30% v/v)
mixtures at a constant mixture velocity of 0.02 m s-1, and flow rate ratio (QIL/QHNO3) 0.5. A- [C4mim][NTf2], B-
[C10mim][NTf2], C- [P66614][NTf2]
To better visualise the circulation patterns, streamlines of the relative velocities are plotted within the plugs. It was found that the continuous phase viscosity affected the circulation patterns. As can be seen in figure 6 the circulation
patterns are more uniform in the case of the less viscous [C4mim][NTf2] (axisymmetric) and become asymmetric when
the more viscous [P66614][NTf2] is used.
Figure 6. Circulation patterns in the aqueous phase for the three different TPB/IL (30% v/v) mixtures at a constant
mixture velocity of 0.02 m s-1, and flow rate ratio (QIL/QHNO3) 0.5. A- [C4mim][NTf2], B- [C10mim][NTf2], C-
0.5 the aqueous phase flow rate increases (since the total volumetric flow rate is constant) and longer plugs are formed. At flow rate ratio 0.5 there are more stagnation points.
Figure 7. Circulation patterns in the aqueous plugs at constant mixture velocity (Vmix) of 0.02 m s-1 with [C4mim][NTf2]
as continuous phase as a function of the flow rate ratio.
4. CONCLUSIONS
The hydrodynamics and the mixing characteristics of a two-phase system [3M nitric acid solution and TBP/IL (30% v/v) mixture] were investigated in a channel of 0.5 mm internal diameter during plug flow with the aqueous phase forming the dispersed plugs. Velocity profiles and film thickness were obtained by means of bright field PIV. It was found that there is a linear decrease of the plug length as the mixture velocity (Vmix) increased when the flow rate ratio
(QIL/QHNO3) is 1. In addition, a thin film was found to be formed between the plugs and the channel wall under all
experimental conditions. The film thickness was found to increase with increasing velocity of the continuous phase (Vc). At a constant mixture velocity it was found that in the cases of the two more viscous ILs ([C10mim][NTf2] and
[P66614][NTf2]) the film thickness was affected by the Vc. In addition, the circulation patterns within plugs became less
uniform with increasing plug length and continuous phase viscosity, while the circulation velocity decreased as the IL viscosity increased. These hydrodynamic characteristics are expected to affect the mass transfer in small channels operating under segmented flow.
REFERENCES
[1] Kashid, M.N., Kiwi-Minsker, L., Quantitative prediction of flow patterns in liquid-liquid flow in microcapillaries. Chem. Eng. Proc., 50, 972-978 (2011)
[2] Kreutzer MT, Kapteijn F, Moulijin JA, and Heiszwolf JJ (2005) Multiphase monolith reactors: chemical reaction engineering of segmented flow in microchannels. Chem. Eng. Sci. 60: 5895-5916.
[3] Wegmann A, and von Rohr PR (2006) Two Phase Liquid–Liquid Flows in Pipes of Small Diameters. International Journal of Multiphase Flow 32 (8): 1017-28.
[4] Kashid MN, Madhvanand N, and Agar DW (2007) Hydrodynamics of Liquid–Liquid Slug Flow Capillary Microreactor: Flow Regimes, Slug Size and Pressure Drop. Chem. Eng. J. 131: 1-13.
[5] Angeli P, and Gavriilidis A (2008) Hydrodynamics of Taylor Flow in Small Channels: A Review. Proc. of the Institution of Mechanical Engineers Part C-Journal of Mechanical Engineering Science 222 (5): 737-51.
[6] Jovanovic J, Zhou W, Rebrov EV, Nijhuis TA, Hessel V, and Schouten JC (2011) Liquid-liquid slug flow: Hydrodynamics and pressure drop. Chem. Eng. Sci. 66: 42-54.
[7] Ghaini A., Mescher A., Agar D.W. (2011). Hydrodynamic studies of liquid-liquid slug flows in circular microchannels. Chem. Eng. Sci. 66, 1168-1178.
[8] Lindken, R., Rossi, M., Große, S., Westerweel, J., Micro-Particle Image Velocimetry (Micropiv): Recent Developments, Applications, and Guidelines. Lab Chip 9 (17), 2551-67 (2009)
[9] Dore, V., Tsaoulidis, D., Angeli, P., Mixing patterns in water plugs during water/ionic liquid segmented flow in microchannels. Chem. Eng. Sci. 80, 334-341 (2012)
[10] Fang WF, Ting SC, Hsu CW, Chen YT, and Yang JT, 2012. Locally enhanced concentration and detection of oligonucleotides in a plug-based microfluidic device. Lab Chip 12, 923-931
[11] Thulasidas, T. C., Abraham, M. A., and Cerro, R. L., 1997. Flow Patterns in Liquid Slugs During Bubble-Train Flow inside Capillaries. Chemical Engineering Science 52 (17), 2947-62.
[12] Van Steijn, V., Kreutzer, M.T., and Kleijn, C. R., 2007. μ-Piv Study of the Formation of Segmented Flow in Microfluidic T-Junctions. Chemical Engineering Science 62 (24), 7505-14.
[13] Freemantle, M., An Introduction to Ionic Liquids. RSC Publications: Cambridge, UK (2010)
[14] Plechkova, N.V. & Seddon, K.R., Applications of ionic liquids in the chemical industry. Chem. Soc. Rev. 37, 123-150 (2008)
